Struct tokenizers::pre_tokenizers::byte_level::ByteLevel

#[non_exhaustive]
pub struct ByteLevel {
    pub add_prefix_space: bool,
    pub trim_offsets: bool,
    pub use_regex: bool,
}

Provides all the necessary steps to handle the BPE tokenization at the byte-level. Takes care of all the required processing steps to transform a UTF-8 string as needed before and after the BPE model does its job.

Fields (Non-exhaustive)

Non-exhaustive structs could have additional fields added in future. Therefore, non-exhaustive structs cannot be constructed in external crates using the traditional Struct { .. } syntax; cannot be matched against without a wildcard ..; and struct update syntax will not work.

add_prefix_space: bool

Whether to add a leading space to the first word. This allows to treat the leading word just as any other word.

trim_offsets: bool

Whether the post processing step should trim offsets to avoid including whitespaces.

use_regex: bool

Whether to use the standard GPT2 regex for whitespace splitting Set it to False if you want to use your own splitting.

Implementations

impl ByteLevel

pub fn new(add_prefix_space: bool, trim_offsets: bool, use_regex: bool) -> Self

pub fn alphabet() -> HashSet<char>

pub fn add_prefix_space(self, v: bool) -> Self

pub fn trim_offsets(self, v: bool) -> Self

pub fn use_regex(self, v: bool) -> Self

Trait Implementations

impl Clone for ByteLevel

fn clone(&self) -> ByteLevel

Returns a copy of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for ByteLevel

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Decoder for ByteLevel

As a Decoder, ByteLevel is in charge of converting any byte-level characters to their unicode counterpart, before merging everything back into a single String. This decoder will consume the tokens and merge them in one step to alleviate the fact that single token decoded might be a byte not representable as as String.

fn decode_chain(&self, tokens: Vec<String>) -> Result<Vec<String>>

fn decode(&self, tokens: Vec<String>) -> Result<String>

impl Default for ByteLevel

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for ByteLevel

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl From<ByteLevel> for DecoderWrapper

fn from(from: ByteLevel) -> Self

Converts to this type from the input type.

impl From<ByteLevel> for PostProcessorWrapper

fn from(from: ByteLevel) -> Self

Converts to this type from the input type.

impl From<ByteLevel> for PreTokenizerWrapper

fn from(from: ByteLevel) -> Self

Converts to this type from the input type.

impl PartialEq for ByteLevel

fn eq(&self, other: &ByteLevel) -> bool

This method tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

This method tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl PostProcessor for ByteLevel

As a PostProcessor, ByteLevel is in charge of trimming the offsets if necessary.

fn added_tokens(&self, _is_pair: bool) -> usize

Returns the number of tokens that will be added during the processing step

fn process_encodings( &self, encodings: Vec<Encoding>, _add_special_tokens: bool ) -> Result<Vec<Encoding>>

Process any amount of encodings and returns a series of encoding (might merge them)

fn process( &self, encoding: Encoding, pair_encoding: Option<Encoding>, add_special_tokens: bool ) -> Result<Encoding>

Process both encodings and returns a new merged one

impl PreTokenizer for ByteLevel

As a PreTokenizer, ByteLevel is in charge of transforming all the unicode characters into their byte-level counterpart. It also splits the input according to the configured regex.

fn pre_tokenize(&self, pretokenized: &mut PreTokenizedString) -> Result<()>

impl Serialize for ByteLevel

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl Copy for ByteLevel

impl Eq for ByteLevel

impl StructuralPartialEq for ByteLevel